To be clear, cesarean delivery saves lives and is one of the most important medical advances ever achieved. There are good reasons that women would deliver by cesarean and good explanations for why some hospitals may deliver more babies by cesarean. If a hospital sees patients with higher complication rates that necessitate cesarean delivery, then variations are not only warranted: they would also be a marker of high-quality care.
But cesarean delivery is also major surgery, with potential risks for both mother and baby compared to vaginal delivery. Cesarean birth is associated with higher rates of infection, re-hospitalization after childbirth, longer and more painful recovery, breastfeeding challenges, problems in future pregnancies, and higher rates of respiratory illness in babies.
Those are all good reasons to avoid cesarean delivery if it’s not medically required. And yet, when we looked at cesarean rates among women with lower-risk pregnancies -- those who have not previously had a cesarean delivery and whose current pregnancy is not preterm, breech, or multiples (twins, triplets, etc.). – we found that cesarean rates varied 15-fold across hospitals, from 2 percent to 36 percent. Nothing in the data appeared to explain these differences, including whether the hospital was in a rural area, its size, or whether the hospital was connected to a university.
Given the risks, the idea that women have such variable odds of cesarean delivery in different hospitals should concern women, clinicians, hospitals and policymakers alike.
These differences aren’t about moms who ask for elective cesareans or doctors who perform them because they fear malpractice suits. Both of these causes have been previously shown to explain a very small fraction of cesarean use. Put simply, the “common sense” reasons do not explain the huge variation across hospitals, particularly among lower-risk women. Understanding this variation and its impact on maternal health and costs is urgent in the context of rising U.S. maternal mortality rates, which have doubled over the past 25 years. 
There are also important financial considerations associated with our findings. Cesareans are more expensive than vaginal births. In 2010, private health insurers paid, on average, $12,739 for hospital care for cesareans versus $9,048 for vaginal deliveries – nearly 30 percent more per delivery. Medicaid programs pay less than private insurers, but cesareans are still more costly, at $4,655 for a cesarean vs. $3,347 for a vaginal birth.
Improving outcomes and quality in maternity care requires efforts to reduce unnecessary variability in cesarean rates. There are two ways we can start now.
First, we need to ensure that pregnant women have access to a range of care options. Many women with healthy pregnancies may not have access to evidence-based care that could be beneficial to them, such as midwifery care, continuous labor support (for example, from a trained doula), and licensed birth centers, which are associated with lower cesarean rates. Low-risk pregnancies should be treated like the normal events they are, with complicated pregnancies warranting specialized care. Medicaid programs and private insurers alike can help facilitate access to appropriate care and may even save money by doing so.

Second, pregnant women need information to make informed decisions about their care, and clinicians, hospitals, and insurers need information about the patterns of care among the women they serve. Women should be able to look up unbiased statistics on the use of cesareans at the hospitals they are considering, and clinicians and hospitals should be aware of their use of cesareans. With accurate data, Medicaid programs and private insurers could tailor incentives to reward high-quality care.
Nearly 4 million American families welcome a new baby each year. How those babies are born should not depend on whether their mother gave birth in one hospital vs. another. As taxpayers, we all have a stake in consistent, evidence-based childbirth care.
Kozhimannil is an assistant professor at the University of Minnesota School of Public Health.